Dynamic simulation analysis for optimal sickbed allocation in a hospital

Objective: Quantitative analyses with combination of data on hospital performance and a mathematical modeling were conducted to evaluate optimal allocation of hospital sickbeds. Methods: Statistical analyses were performed to clarify potential factors which might affect length of (hospital) stay (LOS). In addition, model-based analyses by using methods of system dynamics were performed to elucidate optimal allocation of sickbeds in terms of hospital revenue. A hospital unit which consists of 16 sub-acute stage sickbeds and 44 normal sickbeds were modeled. In this model, hospital admission and discharge were modeled as “flows” which vary the number of patients in sub-acute sickbeds or normal sickbeds in chronological order. Annual hospital revenue derived from all hospitalizations was simulated varying the number of sub-acute stage sickbeds from 10 to 28. Results: Statistical analyses showed advanced age of patients, a hospitalization from external institutions, and having two or more complications were significantly associated with prolonged LOS. Also LOS differed significantly depending on the disease groups. Model-based simulations indicated raising the number of sub-acute sickbeds was associated with an increase in the hospital revenue. Conclusion: Despite several technical challenges, a newly developed approach for optimal bed allocation would be helpful for decision makers and applicable to other hospitals